5753PEIR in low light conditions

I am trying to use an indoor multifunction sensor 5753PEIR (SENPILL having f/w version2.0.00) to turn on light for 3 mins when movement is detected, but only if the ambient light is below a certain level. The ?night move? function for the sensor in Toolkit (version 1.11.6) seems to work OK. However, the sensor is ceiling mounted, pointing down at a darkish coloured floor and the location is often dull but not dark enough to need to turn on the lights. I note earlier posts regarding correct positioning of the units (but it?s too late now) and accurate calibration is not a concern. The light level sensor does not seem to sense lower ambient light levels as being other than zero. The following have been attempted to try to set the ambient light threshold point lower (below which I want the lights to turn on):
Adjust the sensor target light level setting in Toolkit to 10 lux with the hope that the unit potentiometer ?A? then allows more control over the threshold setting.
Left the margin at 9% (changing it to 0% seemed to stop the unit turning on the light at all during movement).
Extensive adjustment of the potentiometer A (toolkit did detect the changes in potentiometer settings). However, it was as though the sensor thought that dull equated to 0 lux according to the ambient light level box in the light level tab. Illuminating the area did cause the reading to go up, so the light level sensor seems to be working OK.
Does anyone have any advice for improving the situation?
Otherwise another approach may be better: there is a PAC available in the system. Would it be feasible to use the indoor sensor as a movement detector and the sunset / sunrise condition of the PAC? There is also an outdoor movement sensor model 5750WPL. Could the light level detected by the outside sensor be combined with movement detected by an indoor sensor, and this used e.g. by the PAC?
NB: there is a 5753PEIR f/w version 2.3.00 in another location of the building ? are they any better at lower light level measurements?

 

The granularity of the threshold setting is in 10 lux steps, so setting the threshold to 10 lux and then trying to further adjust it with the potentiometer will only give you a threshold of either 10 lux or 0 lux. You're better off disabling the potentiometer and setting the threshold in Toolkit.

The sensing performance is the same, so I don't think you'll get any benefit from swapping them. You've got nothing to lose by swapping them over though, so it could be worth a shot.

For low light level applications, I have the target set to 20 lux and the margin set to 0. In fact, the sensor always uses a margin of 10 lux so that there is some hysteresis in the response. This will give you the lowest possible light level threshold for reliable operation. With this configuration, if the sensor is still turning the light on when there's too much light, you'll need to use another method.

There are a bunch of different ways of achieving what you want, all using your existing hardware.

One way would be to set the light level threshold to a higher value, so that the sensor is in effect always operating, and just enable/disable the sensor as required from the PAC. This could be done on time of day, sunrise/sunset or the light level taken from another sensor.

Another way would be to leave the sensor always working, but controlling a dummy group, and then map this group to the real group using the PAC, enabling and disabling the logic module that does this within the PAC as necessary.

 

Thank you Newman.

I had incorrectly assumed that the potentiometer A light level adjustment was somehow an 'analog' stage before the light level measurement was assessed.

Knowing that the granularity setting is in 10 lux stages will save hours more time through not trying to fine tune the potentiomieter. Also, thanks for the guidance on preferred settings and the fact that there is always a 10 lux margin.

For the two alternative solutions mentioned, is there any benefit of choosing one over the other i.e. enabling / disabling the sensor according to sunrise / sunset or enabling / disabling a dummy group?

As for the suggestion of setting the sensor to a highter threshold so that the sensor is in effect always operating, is this better than complementary night move and day move?

For turning off the lights after a predetermined time of detected movement, is it better to use a timer in the PAC or a timer expiry in the sensor? I recall seeing some threads asking whether a sensor can be disabled while the timer is running and expiry of the timer still occurs and turns off the light (but now can not find the answer).

Finally, what is the command / function / technique for reading light levels and using them in a PAC from the outside 5750WPL (just a few keywords is fine for now so I can look this up).

There are a few specific questions here, but any input really is appreciated.

 

The simplest way of doing this is to leave the sensor controlling the light directly, and just enable/disable the sensor remotely from the PAC. All you need to do in Toolkit to the sensor for this is to give it a unique group address for the Enable/Disable on the Occupancy tab. You wouldn't even need logic for this as you could just configure a couple of schedules in the PAC, one to enable the sensor at the right time of day and another to disable it again. Sunset and Sunrise would be good starting points. The Schedules are quite sophisticated in that you can tweak these times as desired, so that you can have events occur at times like "Sunset - 45 minutes". This is the method I would use to start with.

Having the sensor control a dummy group is also fine, but there may be a couple of edge cases to handle to avoid the light getting stranded on if things get disabled whilst the light is on.

It's just a simple method of making the sensor always control it's group address, regardless of light level. You could achieve the same thing by reconfiguring the sensor so that the unit's responses to motion above and below the light level threshold are identical.

I would definitely use the timer in the sensor. This will ensure that the timer is properly re-triggered when motion is re-detected during the timer run time. You can use a PAC timer, but you'll still need to run a really short timer in the sensor and you'll be generating more network traffic than is necesssary.

In PICED you need to set up a Monitor. You can configure this to fetch the current light level from the sensor at regular intervals.